In this paper, I investigate whether moderators adhere to specific, repetitive objectives during guided interaction with interactive digital testimonies and what linguistic practices they use to achieve these goals. The theoretical framework follows Fiehler’s proposition that linguistic, human interaction serves purposes and that these purposes, particularly in an institutional context, can be mapped by specific task hierarchies, from which so-called action schemas (repetitive linguistic action patterns) can be derived (Fiehler 1990, 2016). The empirical data basis for this issue is formed by multiple audio-visual interactional data involving participants from different age groups and a moderator at the Holocaust & Human Rights Museum in Dallas, which were collected in June 2022. Additionally, a questionnaire-guided interview was conducted to supplement the data analysis.